Design Of User Online Behavior Analysis System Based On Network Traffic Classification

In recent years, due to the rapid development of the Internet and the rapid spread of mobile intelligent terminal,the mobile Internet combined with these two has entered a new stage of developments. Nowadays, the mobile Internet has become an important platform for the public to access to the Internet to obtain necessary information, which can be used in convenient browsing, online shopping, chat and other activities. And the mobile Internet is more and more favored by the majority of users. Therefore, many public places such as supermarkets, shopping malls, restaurants and other places deploy the wireless networks to meet the needs of users.So, we can often see a lot of users connecting to the wireless network to conduct network activities in public places through their smart phones. Due to the public internet access places have the characteristic of miscellaneous internet users, it is easy for some criminals to spread rumors and create chaos through the public Internet. And then,these will cause some public places network security issue. At the same time,the increasing Internet users in public places are helpful for us to understand the users' real needs and preferences, and this can provide them with a more personalized interest in the recommendation to help them understand the current location features, thereby promoting the users' consumption.Therefore, it is an important subject to analyze the behavior of users in public places.In this paper, the network traffic generated by Internet users in public places is classified to analyze the Internet users in public places.The main works of this paper are as follows:First of all,to determine the process of network traffic classification and the traffic classification method performance evaluation index by analyzing the network traffic classification method.Secondly, by analyzing the flow characteristics, a network traffic classification method based on data stream fingerprint is proposed.The experimental results show that the proposed method has good classification accuracy and the characteristic of real-time.Finally, the analysis system of users' online behavior in public places is designed.Then the function of users' access to the identity authentication, the guardian process, data transmission protocol and the use of the database are introduced.Users' access authentication function and the classification accuracy rate based on the method of data flow fingerprint network traffic classification are verified.The results show thatthe classification accuracy of the method can meet the practical application requirements.